Adaptive Water Demand Forecasting for Near Real-time Management of Smart Water Distribution Systems United Utilities, a leading UK Water Company, in conjunction with Academia (University of Exeter), have developed a novel methodology to perform short-term Water Demand Forecasting (WDF) with the aim to support near real-time operational management of smart Water Distribution Systems. This data-driven methodology makes use of Evolutionary Artificial Neural Networks. United Utilities is interested in how these type of advanced analytics can be more readily developed and applied i.e. in an agile way, in particular by linking operational data and analytics platforms. With this in mind, we have teamed up with OSIsoft to develop a fully automated Demand Forecasting System (DFS) that exploits OSIsoft AF as well as the PI Integrator for Business Analytics to prepare and bring the near real-time water demand data from PI System to the MATLAB machine learning model. The use of this DFS can assist providing timely operational decisions and supporting tasks, such as production planning, pump scheduling, etc. The results of this proof-of-concept will be presented along with a discussion on the next steps.
Speaker
Kleanthis Mazarakis
Kleanthis graduated from National Technical University of Athens (department of Chemical Engineering) in 2009, and then moved to London for his master degree in Environmental Engineering at Imperial College. He joined OSIsoft in 2012 and since then he has been working in Tech Support and Field service, providing workshops, training and installations, with a specific focus on our water customers as he is Industry Champion for Water Industry.